forked from ipsorakis/GeneralNetworkTools
-
Notifications
You must be signed in to change notification settings - Fork 1
/
sample_random_graph.m
108 lines (95 loc) · 4.22 KB
/
sample_random_graph.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
function [Modularities_per_sample VD_errors] = sample_random_graph(number_of_samplings,max_steps,p,method,datafile,output_filename)
%% INITIALIZE
close all;
load(datafile);
if exist('groups','var')
%VD_errors = zeros(number_of_samplings,length(groups));
Ierrors = zeros(1,number_of_samplings);
end
N = size(W,1);
%% DO A RUN WITH FULL GRAPH
disp('--------RUNNING WITH FULL GRAPH---------------')
if strcmp(method,'eo')
[current_groups current_mod best_iteration best_group] = extremal_optimization(W,max_steps,true,false);
current_mod(end)=[];
elseif strcmp(method,'ng')
[current_groups current_mod best_iteration best_group] = newman_girvan(W,max_steps,true,false);
elseif strcmp(method,'sp')
[current_groups current_mod best_iteration best_group] = spectral_partitioning(W,max_steps,true,false);
elseif strcmp(method,'donetti_munoz')
% DONETTI MUNOZ
[current_groups current_mod best_iteration best_group] = donetti_munoz(W,max_steps,false,'complete-linkage',ceil(N/4),'angular');
elseif strcmp(method,'capocci')
% CAPOCCI
[current_groups current_mod best_iteration best_group] = hierachical_clustering(W,max_steps,false,'complete-linkage',...
'capocci');
elseif strcmp(method,'hierarchical')
% HIERARCHICAL CLUSTERING USING PEARSON SIMILARITY
[current_groups current_mod best_iteration best_group] = hierachical_clustering(W,max_steps,false,'complete-linkage',...
'pearson');
end
Q_full = current_mod
max_steps = max(size(current_mod));
Modularities_per_sample = zeros(number_of_samplings,max_steps);
if exist('groups','var')
Ierror_full = get_normalized_mutual_information(groups,best_group);
end
disp('-----------------------------')
%% SAMPLE
disp('--------BEGINING SAMPLING---------------')
for sample=1:number_of_samplings
sample
Ws = sample_weight_matrix(W,p);
if strcmp(method,'eo')
% EXTREMAL OPTIMIZATION
[current_groups current_mod best_iteration best_group] = extremal_optimization(Ws,max_steps,true,false);
current_mod(end)=[];
elseif strcmp(method,'ng')
% NEWMAN GIRVAN
[current_groups current_mod best_iteration best_group] = newman_girvan(Ws,max_steps,true,false);
elseif strcmp(method,'sp')
% SPECTRAL PARTITIONING
[current_groups current_mod best_iteration best_group] = spectral_partitioning(Ws,max_steps,true,false);
elseif strcmp(method,'donetti_munoz')
% DONETTI MUNOZ
[current_groups current_mod best_iteration best_group] = donetti_munoz(Ws,max_steps,false,'complete-linkage',ceil(N/4),'angular');
elseif strcmp(method,'capocci')
% CAPOCCI
[current_groups current_mod best_iteration best_group] = hierachical_clustering(Ws,max_steps,false,'complete-linkage',...
'capocci');
elseif strcmp(method,'hierarchical')
% HIERARCHICAL CLUSTERING USING PEARSON SIMILARITY
[current_groups current_mod best_iteration best_group] = hierachical_clustering(Ws,max_steps,false,'complete-linkage',...
'pearson');
end
Modularities_per_sample(sample,:) = current_mod;
current_mod
if exist('groups','var')
%[C d VD_error] = partitioning_error(groups,best_group);
%VD_errors(sample,:)=VD_error';
Ierrors(sample) = get_normalized_mutual_information(groups,best_group);
end
disp('-----------------------------')
end
% if exist('groups','var')
% mean_VD_errors = mean(VD_error);
% end
%% DO SOME PLOTTING
meanplot = mean(Modularities_per_sample,1);
stdplot = std(Modularities_per_sample,1);
hold on
myfig = errorbar(meanplot,stdplot);
plot(Q_full,'--r','LineWidth',2)
hline(get_modularity(groups,W),'g','observed Q');
title(strcat('Modularity across algorithm steps, ',method,', p=',num2str(p),', dataset=',datafile));
ylabel('Modularity Q');
xlabel('Algorithm steps (dendogram layers)');
legend(strcat(num2str(number_of_samplings),' samples'),'full graph');
hold off
%% SAVE IF NEEDED
if exist('output_filename','var')
clear current_groups;
save(strcat(output_filename,'_p',num2str(p*100),'_',method));
saveas(myfig,strcat(output_filename,'_p',num2str(p*100),'_',method));
end
end